Skip to main content

Data refreshed daily · DR by Ahrefs

Sign in

Backlink Quality Versus Quantity: Reading a Link Profile

Links · 10 min read ·

Not all links are equal, and counting them is not judging them. How to tell a healthy link profile from a padded one by relevance, source and context.

Illustration: A scatter-style chart on light grey with points plotted by relevance and audience, a cluster of strong links top right in teal and a pile of weak ones bottom left in amber

A link report offers a flattering number: how many sites link to you. It is easy to read it as a scoreboard. More is better, so more must be the goal.

It is not that simple. Links vary enormously in what they are worth, and some are worth less than nothing. This guide is about reading a link profile with judgement: what quality means, what to look for, how to tell padding from substance and why chasing quantity is a trap.

A backlink is a link from one site to another. Search engines have long used links as one signal that a page is worth attention, on the logic that people link to things they find useful. Link-analysis companies measure that signal. Ahrefs' Domain Rating, for example, estimates the relative strength of a site's backlink profile, and it counts followed links and only the first link from each referring domain.

Two facts follow from how it works.

First, quantity of links from one site does not add up. Ten links from the same domain count, for DR purposes, as one referring domain. The measure rewards diversity of sources.

Second, the strength of each source matters. A link from a respected, widely linked site passes more than a link from an obscure one.

But DR is a score of links, not a verdict on whether the links are good. For that, you must look.

Think of a link as a recommendation. A good recommendation comes from someone credible, in context, for a reason. Applied to links, that suggests five tests.

Relevance

Does the linking page relate to your subject? A link to a note-taking app from a page about note-taking tools makes sense. A link from an unrelated gambling site does not. Relevance is the clearest sign of a natural link.

A real audience

Does the linking page have real readers? A page that nobody reads cannot send visitors, and it suggests the link was not placed for people.

Editorial choice

Did a person choose to link, because they found your page useful? Links given freely are the gold standard. Links that were paid for, swapped or placed in bulk are different.

Natural context

Is the link in the body of useful content, with sensible surrounding words? Or in a list of dozens of unrelated links, a footer, a signature or a block of spam?

Stability

Is the linking page likely to exist next year? Pages on established sites tend to last. Pages made quickly for the purpose of linking often vanish.

Score a link on these five and you will have a rough but honest sense of its worth.

Warning signs:

  • Irrelevant source. Nothing in common with your subject.
  • Link farms and pages built to hold links.
  • Mass directory submissions with no editorial review.
  • Paid links that are not marked. Search engines expect paid or sponsored links to be qualified with the right attribute.
  • Reciprocal link schemes at scale.
  • Comment and forum spam.
  • Unnatural anchor text repeated in the same form across many sites.
  • Sudden bursts from unrelated sources.
  • Foreign-language or auto-generated pages with no connection to you.

Google's spam policies describe link spam as links intended to manipulate ranking, including buying or selling links that pass ranking credit, excessive link exchanges and automated link creation. Its guidance on qualifying outbound links explains the nofollow, sponsored and user-generated attributes that mark such links so they do not pass credit. If a link profile is full of these patterns, it may do more harm than good.

Why quantity misleads

There are several reasons why raw counts mislead.

Counts hide quality. Ten thousand weak links are worth less than ten strong ones.

Counts hide concentration. Thousands of links from one domain look impressive and count as one source.

Counts reward manipulation. It is cheap to make many links. Because counts are easy to inflate, they are easy to fake.

Counts reward the wrong behaviour. If you aim at a number, you buy volume, not value.

Ahrefs states that DR does not account for backlink spam and that large amounts of low-quality backlinks may increase DR. A score can therefore rise because of the very links that a search engine would distrust. A healthy profile is judged by looking inside it.

How to read your own profile

You can do this in an afternoon.

  1. Export the list of referring domains, not just links. This gives you diversity at a glance.
  2. Sort by strength, whatever measure your tool uses, and look at the top twenty. Are they sensible, relevant, real sites?
  3. Sort by newest, and look at recent arrivals. Is there a natural flow, or an odd cluster?
  4. Look at the anchor text: the words in the link. A natural profile has a mix: your name, plain phrases, page titles. A profile where most anchors are the same commercial phrase is suspicious.
  5. Look at the pages linking to you, not just the sites. Open a sample. Are they real?
  6. Look at the ratio of followed to nofollow. Many legitimate profiles contain a lot of nofollow links, and that is normal.
  7. Look at the countries and languages, if your audience is specific. Large unexplained foreign clusters can be a flag.
  8. Look for lost links. Which valuable ones have gone?

Write down three observations: the best links, the worst and the pattern.

Worked example in words

A small software site has a few dozen referring domains. The top five are a well-known newsletter, a community forum thread, a directory listing with editorial review, a blog review by an independent writer and a university resource page. Each is relevant, has real readers and was given freely. The next twenty are a mix of smaller blogs and directories. Four are low-quality scraper sites.

Compare with another site with thousands of links, nearly all from a handful of auto-generated pages with the same anchor phrase. The second site's report shows a bigger number and a weaker position. If the first site wants to improve, the lesson is to earn more links like the top five, not to pursue volume.

If you find poor links pointing to you:

  • Do not panic. Search engines are generally good at ignoring junk, and every large site attracts some.
  • Do not buy cleaning services that promise miracles.
  • Look for patterns you caused: old paid arrangements, directories you submitted to carelessly, an agency's past work.
  • Ask to remove links you caused, by contacting the site owner politely.
  • Use disavow tools only with care and advice. Search engines offer a way to tell them to ignore certain links, but it is a blunt instrument and rarely needed.
  • Focus on earning good links rather than fighting every bad one.

On this platform, spam checks are part of how followed links are decided. A followed link depends on the listed site meeting a DR threshold set by our team and passing those checks, with a disavow export available, and free listings are marked nofollow. The aim is to keep the profile a link from here sits in clean.

Quality links follow quality work.

  • Make something worth linking to: a useful tool, an honest guide, original data, a clear explanation.
  • Be present where your audience gathers, and be helpful.
  • Get listed in places with real editorial standards.
  • Contribute to the community with talks, answers and open source.
  • Reach out sparingly, with something genuinely relevant, to people who might find it useful.
  • Publish things people cite: a well-made reference page, a clear comparison, a dataset.
  • Say thank you to people who link, and keep their pages accurate.

None of this is fast. It is also the only approach that is hard to copy and unlikely to backfire.

The mix to expect

A natural profile is varied.

  • Some followed, some nofollow, some sponsored where paid.
  • Links to your home page and to deeper pages.
  • A range of anchor words.
  • A range of site sizes.
  • A flow of new links and some natural loss.

Do not aim for perfection. Aim for plausibility: a profile that looks like what a genuinely useful site would attract.

DR in context

The live DR shown here is currently 0. That is a number from a link index, with a check date and a signed change. It is a useful tile and a poor target. Read it with the quality checks above: a rising DR driven by a few good new referring domains is a good sign, and a rising DR driven by volume from one source is not.

Common mistakes

  • Counting links, not domains.
  • Buying packages.
  • Ignoring relevance.
  • Over-optimising anchor text.
  • Chasing every low-DR site or only high-DR ones. Relevance matters more than score.
  • Treating nofollow as worthless. It still brings visitors and credibility.
  • Never auditing.

On this site

The leaderboards and launches pages show DR with its trend and check dates, and the categories page helps you find where relevant neighbours sit. The pricing page explains what is free and optional, the developers page describes the free API and MCP server and the submit page starts a listing.

The short version

Count referring domains, not raw links, and judge each link on relevance, real audience, editorial choice, natural context and stability. Be wary of volume, paid links that are not marked and anchor patterns that look manufactured. Audit your profile, do not panic over junk, never buy bulk and earn links by making something worth linking to. A few good recommendations beat a crowd of bad ones.

Questions and answers

Is it better to have many links or good links?
Good links. A few relevant, editorially given links from real sites generally outweigh many low-quality or manufactured ones.
What makes a link good?
Relevance, a real audience on the linking page, editorial choice, a natural context and a source that would exist without the link.
How do I recognise a poor link?
Signs include irrelevant sites, pages made only to hold links, paid arrangements that are not marked, mass directories and sudden unnatural patterns.
Does the number of referring domains matter?
It is a useful measure of diversity, since only the first link from a domain raises DR, but quality still matters more.
What is this site's DR?
The live figure is 0, shown publicly with its check date and change.

Sources

Ask us